On Mon, 2008-05-05 at 16:05 +0800, Zhu Yi wrote: > On Mon, 2008-05-05 at 09:11 +0200, Johannes Berg wrote: > > Those conditions will never be true. > > Why? I see mac80211 never passes IEEE80211_IF_TYPE_MNTR to > add_interface(). But this patch passes the flag in configure_filter(). > Please take a look. That's not a flag. Yes, I did miss that call iwl4965_set_mode(priv, IEEE80211_IF_TYPE_MNTR) but I still don't think you should treat monitors as a "mode" but rather as a reason to pass more frames up. > > In the case of the latter, that is > > a good thing, in the case of the former that isn't. > > Could you elaborate? You're precluding doing any kind of useful monitor operation when the card is actually up&running, which is quite useful for debugging. By the second condition, you're short-cutting the RX path so that when a monitor is enabled any other type of interface won't operation properly, as far as I can tell. johannes
Attachment:
signature.asc
Description: This is a digitally signed message part
